Few-shot object detection on aerial imagery via deep metric learning and knowledge inheritance

Li, Wu-zhou; Zhou, Jian; Li, Xiang; Cao, Yi; Jin, Guang

doi:10.1016/j.jag.2023.103397

Cited by 8 publications

(4 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Various triplet selection strategies, like semi-hard triplet mining, as used on FaceNet [29], aim to strike a balance between excessively challenging and overly simplistic examples, located within a predefined margin hyperparameter. DML has been applied to tasks such as face classification, hyperspectral image classification, and few-shot classifications of diseases [25,[29][30][31][32][33][34], where it has outperformed the baseline methods.…”

Section: Introductionmentioning

confidence: 99%

Pine Wilt Disease Segmentation with Deep Metric Learning Species Classification for Early-Stage Disease and Potential False Positive Identification

Thapa,

Khanal,

Bhattarai

et al. 2024

Electronics

View full text Add to dashboard Cite

Pine Wilt Disease poses a significant global threat to forests, necessitating swift detection methods. Conventional approaches are resource-intensive but utilizing deep learning on ortho-mapped images obtained from Unmanned Aerial Vehicles offers cost-effective and scalable solutions. This study presents a novel method for Pine Wilt Disease detection and classification using YOLOv8 for segmenting diseased areas, followed by cropping the diseased regions from the original image and applying Deep Metric Learning for classification. We trained a ResNet50 model using semi-hard triplet loss to obtain embeddings, and subsequently trained a Random Forest classifier tasked with identifying tree species and distinguishing false positives. Segmentation was favored over object detection due to its ability to provide pixel-level information, enabling the flexible extension of subsequent bounding boxes. Deep Metric Learning-based classification after segmentation was chosen for its effectiveness in handling visually similar images. The results indicate a mean Intersection over Union of 83.12% for segmentation, with classification accuracies of 98.7% and 90.7% on the validation and test sets, respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

Pine Wilt Disease Segmentation with Deep Metric Learning Species Classification for Early-Stage Disease and Potential False Positive Identification

Thapa,

Khanal,

Bhattarai

et al. 2024

Electronics

View full text Add to dashboard Cite

show abstract

“…In recent years, we have been witnessing a rapid development in Very-High-Resolution Satellite imaging (VHRS). It may be applied in numerous fields: landcover mapping [1,2], urban mapping [3], the detection and tracking of objects [4][5][6][7], maritime monitoring [8,9], automatic building classification, etc., [10,11].…”

Section: Introductionmentioning

confidence: 99%

Modified ESRGAN with Uformer for Video Satellite Imagery Super-Resolution

Karwowska,

Wierzbicki

2024

Remote Sensing

View full text Add to dashboard Cite

In recent years, a growing number of sensors that provide imagery with constantly increasing spatial resolution are being placed on the orbit. Contemporary Very-High-Resolution Satellites (VHRS) are capable of recording images with a spatial resolution of less than 0.30 m. However, until now, these scenes were acquired in a static way. The new technique of the dynamic acquisition of video satellite imagery has been available only for a few years. It has multiple applications related to remote sensing. However, in spite of the offered possibility to detect dynamic targets, its main limitation is the degradation of the spatial resolution of the image that results from imaging in video mode, along with a significant influence of lossy compression. This article presents a methodology that employs Generative Adversarial Networks (GAN). For this purpose, a modified ESRGAN architecture is used for the spatial resolution enhancement of video satellite images. In this solution, the GAN network generator was extended by the Uformer model, which is responsible for a significant improvement in the quality of the estimated SR images. This enhances the possibilities to recognize and detect objects significantly. The discussed solution was tested on the Jilin-1 dataset and it presents the best results for both the global and local assessment of the image (the mean values of the SSIM and PSNR parameters for the test data were, respectively, 0.98 and 38.32 dB). Additionally, the proposed solution, in spite of the fact that it employs artificial neural networks, does not require a high computational capacity, which means it can be implemented in workstations that are not equipped with graphic processors.

show abstract

“…Among the several different DL architectures, such as deep belief networks [27,28], generative adversarial networks [29], transformers [30] and autoencoders [31,32], convolutional neural networks (CNNs) and recurrent neural networks (RNNs) are the most extensively used models for complex remote sensing applications [33,34]. Prominent examples are satellite image fusion for improved land use/land cover classification [35], object detection [36,37], and change detection [38,39] in remote sensing images, as well as the delineation of agricultural fields from satellite images [40]. CNNs are appealing to the remote sensing community due to their inherent nature to exploit the two-dimensional structure of images, efficiently extracting spectral and spatial features, while RNNs can handle sequential input in continuous dimensions with sequential long-range dependency, thus making them appropriate for the analysis of the spectral-temporal information in time series stacks [19,34,[41][42][43][44].…”

Section: Introductionmentioning

confidence: 99%

Agricultural Land Cover Mapping through Two Deep Learning Models in the Framework of EU’s CAP Activities Using Sentinel-2 Multitemporal Imagery

Papadopoulou,

Mallinis,

Siachalou

et al. 2023

Remote Sensing

View full text Add to dashboard Cite

The images of the Sentinel-2 constellation can help the verification process of farmers’ declarations, providing, among other things, accurate spatial explicit maps of the agricultural land cover. The aim of the study is to design, develop, and evaluate two deep learning (DL) architectures tailored for agricultural land cover and crop type mapping. The focus is on a detailed class scheme encompassing fifteen distinct classes, utilizing Sentinel-2 imagery acquired on a monthly basis throughout the year. The study’s geographical scope covers a diverse rural area in North Greece, situated within southeast Europe. These architectures are a Temporal Convolutional Neural Network (CNN) and a combination of a Recurrent and a 2D Convolutional Neural Network (R-CNN), and their accuracy is compared to the well-established Random Forest (RF) machine learning algorithm. The comparative approach is not restricted to simply presenting the results given by classification metrics, but it also assesses the uncertainty of the classification results using an entropy measure and the spatial distribution of the classification errors. Furthermore, the issue of sampling strategy for the extraction of the training set is highlighted, targeting the efficient handling of both the imbalance of the dataset and the spectral variability of instances among classes. The two developed deep learning architectures performed equally well, presenting an overall accuracy of 90.13% (Temporal CNN) and 90.18% (R-CNN), higher than the 86.31% overall accuracy of the RF approach. Finally, the Temporal CNN method presented a lower entropy value (6.63%), compared both to R-CNN (7.76%) and RF (28.94%) methods, indicating that both DL approaches should be considered for developing operational EO processing workflows.

show abstract

Few-shot object detection on aerial imagery via deep metric learning and knowledge inheritance

Cited by 8 publications

References 35 publications

Pine Wilt Disease Segmentation with Deep Metric Learning Species Classification for Early-Stage Disease and Potential False Positive Identification

Pine Wilt Disease Segmentation with Deep Metric Learning Species Classification for Early-Stage Disease and Potential False Positive Identification

Modified ESRGAN with Uformer for Video Satellite Imagery Super-Resolution

Agricultural Land Cover Mapping through Two Deep Learning Models in the Framework of EU’s CAP Activities Using Sentinel-2 Multitemporal Imagery

Contact Info

Product

Resources

About